Hacker News Reader

The eHang 184 looks like a device to decapitate many people quickly, or capping kneecaps while on the ground.

The Volocopter (with 18 rotors above instead of 8 below) looks like a much more reasonable and fault tolerant concept (that has actually had manned flights for years).

EDIT to add: lazy journalists - "World’s First Passenger Drone Unveiled In China" in 2018, while the Volocopter had its first manned flight in 2016 [1] and been flying around in Dubai 2017 [2]... Ignorance is excusable, but not doing even a cursory internet check isn't really.

[1] https://www.youtube.com/watch?v=OazFiIhwAEs

[2] https://www.youtube.com/watch?v=ROJ76foyihs

Propeller airplanes are not different, in fact real accidents happen with these airplanes. This didn't stop such airplanes from being widely used. Given a well controlled situation, vehicles such as the eHang can very well be used.

I am horrified by eHang's propellers. Can't they add some sort of guard around them or move them above the cabin? I can't stop thinking that even passengers would run into them frequently when stepping on or off the thing, not to mention loss of control situations..

Well yes, and airplanes only operate on airfields with generally carefully controlled access, and still those accidents happen. Now, with airplanes there is really no other option, but with a drone one can easily place the rotors above, rather than below, the passenger cabin (which also gives better stability).

The eHang 184 is truly a horrendous design in my opinion, and I am puzzled why it is seriously reported about.

> one can easily place the rotors above […] (which also gives better stability).

https://en.wikipedia.org/wiki/Pendulum_rocket_fallacy

Very interesting. I’m not sure whether that fallacy applies.

Hard to find authoritative statements on it. There is a thread on aviationexchange where some say that an underslung helicopter is actually more stable than a conventional one, but I haven’t been able to verify this.

With a multicopter, the situation might be different still.

> Someone recently flew a jet-powered "hoverboard" across the English channel

And stopped for a refuel half-way.

Still looks surprisingly useful even in its current form.

Yeah. If the thing is not at least in an X8 configuration (2 independent motors and props for each edge) with enough thrust to safely land you in X4, it should never have passengers on board.

The car shown in OP is one motor/esc/prop failure away from killing whoever is inside it.

And whoever happens to be below it at the time.

Couldn’t the same be said about any helicopter? It’s a single blade. A single failure mode. And they routinely ferry tourists and non-professionals.

Helicopters utilize autorotation for emergency landing.

>That has four sets of contra-rotating props, so there's some redundancy

That is no redundancy. With only four booms, it only takes a single boom failure (electrical or mechanical) to crash. And with co-axial counter-rotating props, a prop failure will almost certainly cascade to its partner - so again, you really only have four props.

There is almost no survivable failure mode with that design.

>There is almost no survivable failure mode with that design.

Parachute?

Parachutes can help, but even with rocket assist there can be a "dead zone": too high for passengers to survive a crash but too low for the parachute to fully deploy. The aircraft also has to be fairly stable for the parachute to deploy successfully; if the aircraft has a sharp pitch or roll speed due to thrust imbalance then the parachute might not work.

Maybe a modified version of this skyscraper escape system https://i.redd.it/5ltd9qizlh621.gif ?

There should be ways to deploy parachutes faster. Maybe with springs or pressurized gas.

Otherwise having large airbags deploy under the vehicle could be another option.

If just one wheel fails on a car travelling at 100km/h on a road with oncoming traffic you're very unlikely to survive either.

A 200km/h head on collision with equivalent modern cars is absolutely survivable. Gravity is pretty potent at getting you moving faster than that, and with extremely uncertain impact point

Terminal velocity for a person is around 195kph.

That’s minimum arms out speeds for a human, you can hit 400+kph in a streamlined dive. Failing aircraft can go well past that into ripping of the wings of an aircraft.

I know we're veering somewhat off topic here, but I wonder how true this is in practice?

I can't find a lot of testing done at this sort of speed with regular cars. Racing cars and some high-end sports cars are exceptionally strong and allow you to survive a 200 km/h crash - often with only very minor injuries - but I'm really uncertain that this is true for a regular family car. Do you have any further data on this? Given how many regular cars I see driven at this sort of speed when I visit Germany, it'd be interesting to know!

It's actually a 100km/h collision done twice. Each car goes from 100 to 0, not from 200 to 0. There's twice the energy but it's distributed across the two cars so for each one it's equivalent to hitting a stationary wall.

That said, based on this video I don't think an actual 200km/h is anywhere near survivable for the average modern car: https://www.youtube.com/watch?v=LmRkPyuet_o

Possibly higher end sports cars with cages and five point harnesses may fare better.

A 100km/h head-on with a similar car going 100km/h should be similar to a 100km/h head-on collision with a solid concrete wall. However a car traveling at 200km/h has four times the kinetic energy as a car traveling at 100km/h so a collision at that speed is much less likely to be survived than a collision at 100km/h.

Wow you are the first person to point out what I’ve been saying for decades. A car hitting another car at the same speed is indistinguishable from a car hitting an immovable wall.

Intuition wants to think it’s twice as bad.

As an aside can I pose another physics car challenge to you that 100% of people I’ve asked get wrong?

You’re the only car at stop light waiting for the light to turn green. You look up in your rear view mirror and see a car bearing down on you presumably with a distracted driver. You know you’re about to get rear-ended. In that split second do you: A) Let up on the brakes and put the car in Neutral or B) slam in the brake as hard as you can.

The brake will increase survivability for you; neutral will increase survivability for the other driver. What hurts you is the sudden acceleration from the impact, and the brake will decrease that. What hurts the other driver is the sudden deceleration of impact; your brake will make that a harder stop for him while being in neutral would make it a bit less of a hard stop.

This is backwards.

Your car will accelerate suddenly from the impact either way, so you want to increase the time-frame over which the impact forces of the crash are dissipated. If your car is in neutral, then the crash is less severe because only part of the other car hits you before your car starts gets pushed forward, so more of the energy of the impact gets transmuted to accelerating your car and less energy needs to be absorbed by your car's crumple zones.

If you slam on the brakes, you effectively turn your car into a stationary object, and that means you and the other car must absorb the full energy of the collision. At collision speeds where survivability is a concern, this usually means the crumple zones are inadequate to absorb the energy of the collision and so more of it gets passed on to your body.

Ideally, you would want to accelerate your car as much as possible before the impact to minimize the speed differential. Even a 10mph decrease in the differential has a huge impact on survivability.

[Note: bigger cars can survive crashes better because they have more mass they can act as crumple zones to absorb the force of the crash. Older, more rigid cars would frequently survive with limited damage crashes that killed their occupants, at lower average speeds than today, because the cars transmitted more of the impact forces to the occupants. Modern cars are designed to crumple to dissipate impact forces.]

Surely the answer there depends on what's in the intersection - whether you want to roll into it or not.

The answer may well be c) gun it and hope for the best!

WRT the two cars crashing - I would bear in mind that immovable objects are quite uncommon, most highway barriers are compressible. So whilst you might be strictly correct I'd imagine few people crash into completely stationary objects like bridge supports without crash buffers.

> However a car traveling at 200km/h has four times the kinetic energy as a car traveling at 100km/h so a collision at that speed is much less likely to be survived than a collision at 100km/h.

Kinetic energy by itself does not matter one bit when it comes to how damaging a crash will be for the human occupants. People use it as a proxy for acceleration and jerk. Kinetic energy is only actually useful when you're trying to figure out what acceleration and jerk will be if the crash involves multiple moving vehicles or hitting an object that's light enough to move.

Imagine a freight train crashing into an immovable something at low speed. The kinetic energy is going to be higher than almost any car crash but the occupants will be just fine and they have far less crash protection stuff around them than anyone in a car. Alternatively, imagine an occupied small car traveling at some low speed crashing head on into a person who was shot out of a cannon at high enough speed for their kinetic energy to be equal. Both parties are going to decelerate over the same amount of time but the person shot out of the cannon will experience much higher acceleration (and jerk) because their initial velocity was higher.

The kinetic energy of the human is the important part, and higher speed increases that, unlike higher mass of vehicle as you note.

If it falls off that axle is only dropping 20cm though, instead of however high this thing flies. Plus, wheels are a lot more solid than props.

Wheels are pretty solidly mounted to cars with large pieces of metal. Tires do blow out however then you still have some control.

Props on the other hands can have electrical component failures, wiring failures, engine failure, bird ingestion induced failures, control instrument failures, etc. Remember that each vehicle has four times these components (ie: four times the failure rate) and the failure of any is a crash.

Just a lot more potential things that can fail and cause a catastrophic accident.

Actually you're pretty likely to survive as long as you're wearing a seat-belt and don't crash into anything that only hits the upper half of the cabin (like the side of a big truck)

I'm pretty sure that it is generally considered a more reliable design in both small drones and such vechicles. Yes, there are 4 booms, but there are 8 ESCs, 8 cable sets, 8 motors, 8 motor holders(they are the ones experiencing most vibrations), 8 props.

All those 8x parts are much more likely to fail than the boom. How would the boom even fail mechanically if it's just a solid holding mechanical part that is well-built and tested? Would it just break off? Because of what? If so, in any regular normal airplane or a helicopter there are many other similar parts which are usually not redundant. This must be due to the fact that such solid parts fail very rarely, and perhaps more importantly, can be somewhat easily tested on the ground to check their structural integrity. Motors, ESCs, etc can fail due to many other reasons and are sometimes harder to test properly, there is much more risk in them.

So yes there is redundancy.

This configuration of props is called coax, and is specifically less effective than just putting 4 motors of higher power and bigger props. If there were no gains in reliability, no one would build drones like this, because they would just be less efficient and fly for less time. There is basically no other advantage in this setup than their better reliability due to redundancy.

>And with co-axial counter-rotating props, a prop failure will almost certainly cascade to its partner

Do you have a source for that?

If a prop breaks, it will likely fly off onto sides due to centrifugal forces, not on the vertical axis into the other prop. Additionally if a bottom prop breaks, there is already a downward flow of air from the top prop, which will act as additional downward force on those breakaway parts which will force them to fly off downwards, not upwards into the remaining prop.

My source is bitter experience designing and testing counter-rotating electric ducted fans for distributed propulsion drone applications.

You have to consider specific failure modes. You might have an object strike, like a bird. You might lose sealing and get water in somewhere, frying anything electronic inside. You might have a loose bolt, a fatigued bracket, or a frayed wire. When you start considering scenarios rather than parts lists, it becomes clear that consolidating things together spatially costs you most of the advantage of redundancy.

This is especially true in the case of props. In point of fact, if a prop breaks the net force is upwards, because that's the direction it was pushing in when it failed. However it's basically irrelevant - unless you perfectly and flawlessly detach the blades symmetrically with explosive bolts, with impeccable timing such that no detached blade hits the airframe, so much energy is released with a moving-part failure that anything in the vicinity is pretty well destroyed. Experience you might have with normal-sized drones is not applicable if you scale it up even a little - big props make a big mess, and you're lucky if the airframe survives at all. You certainly won't be able to count on that shared pod.

We can only guess at why they chose to design it this way - personally I think they're a bunch of cowboys who will kill someone the way they're going. I suspect it's because they want to keep the footprint small, and because it's kind of uncool to have to pick your way through a dense ring of props to get to the cockpit (if you look at some of their videos you'll see that they've experimented with non-coaxial x8 configurations in the past). But stacking the props two-by-two like that only costs you reliability.

I would think that the most likely failure mode would be bird ingestion. A large bird would likely break the top prop, and the fragments of that prop+bird would hit the cowel, and then be sucked down through the bottom prop. Not to mention the high forces exerted on the top-prop motor mounting from losing only one blade. I'm not sure whether you want the top prop motor mount to fail at that point to save the airframe, but if so, then you'll get the top motor ingested into the bottom prop too. Unless you can actually prevent large bird ingestion, or design props that can survive it intact, I agree you don't really have any redundancy from co-axial contra-rotating props.

A traditional helicopter had no redundancy either. The single motor could fail or a single blade. What’s the difference?

See https://en.wikipedia.org/wiki/Autorotation

Ballistic parachutes. They've been a safety feature in composite aircraft like the Cirrus SR22 for a long time.

Useless on commercial aircraft, great for small, lightweight personal transportation vehicles.

I'm a complete layman in aeronautics, but Lilium's prototypes seem to look more promising (over-optimistic?) on paper and make me worry less about decapitation.

https://www.youtube.com/watch?v=bQSDKmVoaIo

> Someone recently flew a jet-powered "hoverboard" across the English channel.That thing is powered by several large model aircraft jet engines. It's much like the Hiller Flying Platform of 1954.

Not sure the Zapata "flyboard" is fully consumer-ready. The inventor lost two fingers flying an earlier version. https://www.core77.com/posts/89748/Franky-Zapata-Crossed-the...

That only suggests the earlier version wasn't consumer ready. And people injure themselves with mass produced consumer ready vehicles every day.

A post flight interview suggested that the hoverboard requires very high skill levels and leg fitness.

I personally prefer the Williams X-Jet or "flying pulpit" from the 80s:

https://youtu.be/guc56Kbz5yI

Pity it never went into mass production. We should crowdfund a modern version!

Looks pretty advanced for 1980s tech. The army passed on it because it didn’t perform better than alternatives they already had.

The army might have bought in on this if they had mounted a turret or some offensive weapon system of some sort though, otherwise to them it’s just a toy.

Ground Level Chemtrail Distribution via Flying Saucer!